Re: [RFC V2] mm: Enable generic pfn_valid() to handle early sections with memmap holes

[Mike Rapoport]

On Thu, Apr 22, 2021 at 04:53:36PM +0530, Anshuman Khandual wrote:
> On 4/22/21 1:57 PM, Mike Rapoport wrote:

...

> >> diff --git a/mm/memblock.c b/mm/memblock.c
> >> index 3abf2c3fea7f..93f8a9c8428d 100644
> >> --- a/mm/memblock.c
> >> +++ b/mm/memblock.c
> >> @@ -1740,6 +1740,7 @@ bool __init_memblock memblock_is_memory(phys_addr_t addr)
> >>  {
> >>  	return memblock_search(&memblock.memory, addr) != -1;
> >>  }
> >> +EXPORT_SYMBOL(memblock_is_memory);
> > 
> > Please make it inside #ifdef CONFIG_ARCH_MEMBLOCK
> CONFIG_ARCH_KEEP_MEMBLOCK ?

Yeah, _KEEP went away somehow :)

> Wrap it around the EXPORT_SYMBOL() or the entire function
> memblock_is_memory().

EXPORT_SYMBOL(). Otherwise we'll have exported __init function.
 
-- 
Sincerely yours,
Mike.